12月1月2月3月4月5月6月7月8月9月10月11月
周一周三周五
一年内 263412 次贡献
更少更多
- frostbyte_neo 推送了仓库 github-featured/smol-machines--smolvm 的 fix-pack-dir-modes 分支d1fd592f94 Restore every directory's mode after extracting a pack, not only unwritable ones
- frostbyte_neo 推送了仓库 github-featured/smol-machines--smolvm 的 fix-init-bake-proxy 分支5fdcdafa3b Merge remote-tracking branch 'origin/main' into fix-init-bake-proxy
- frostbyte_neo 推送了仓库 github-featured/smol-machines--smolvm 的 fix-fromvm-flatten 分支比较 4 提交 »b7c33a17c2 Drop an unnecessary mut in the flatten testsca158f1eb9 Stack the flattened layers topmost first so later image layers and machine writes win1bd6813691 Carry the image's environment into from-vm packs so its PATH still resolves77dab73e98 Flatten from-vm packs in the agent so images with four or more layers can be packed
- frostbyte_neo 推送了仓库 github-featured/smol-machines--smolvm 的 egress-denial-events 分支比较 2 提交 »706b5bdd66 Write egress denials to a dedicated audit file so connection chatter cannot evict them0365fbea4c Record egress denials and surface them per machine through the CLI and serve API
- frostbyte_neo 推送了仓库 github-featured/smol-machines--smolvm 的 fix-clone-identity 分支23a59e0492 Rejuvenate a forked clone's identity inside the workload rootfs so image clones stop inheriting the golden's SSH host keys
- frostbyte_neo 推送了仓库 github-featured/smol-machines--smolvm 的 fix-cuda-chunks-lint 分支186166c1a7 Modernize for the clippy 1.98 lints: as_chunks over constant chunks_exact, and bounded loops over unbounded for-ranges
- frostbyte_neo 推送了仓库 github-featured/smol-machines--smolvm 的 cuda-segmented-graph 分支比较 10 提交 »51a5f8dd81 Stabilize forkpoint readiness race tests8b28cdc8cd Update h2 to address empty frame denial of service8fd61b2ceb Accept staged layers in rootless shared-extract testscd82976cb5 Forward CUDA graph updates through runtime and driver APIs3ff0757aec Harden CUDA graph capture and event handling
- frostbyte_neo 推送了仓库 github-featured/smol-machines--smolvm 的 bump-v1.9.0 分支199821fd67 Bump the workspace to 1.9.0
- frostbyte_neo 推送了仓库 github-featured/smol-machines--smolvm 的 cuda-auto-replay 分支比较 6 提交 »c34d6bf3c2 Remove provider details from CUDA README679d1fde39 Keep CUDA capacity portable20c2eac986 Expose CUDA capacity to fleet schedulersb6f844533f Keep automatic CUDA replay safe for distributed workloads159ca37739 Keep mapped host memory Linux-only
- frostbyte_neo 推送了仓库 github-featured/smol-machines--smolvm 的 bump-h2-rustsec-2026-0258 分支dd2319a280 Update h2 to 0.4.16 to clear the RUSTSEC-2026-0258 empty-DATA-frames advisory
